Episode 1 · Visitor at the Bedside · 9:16 Vertical — AI Video from Neta Studio

Episode 1 · Visitor at the Bedside is the vertical cut of the opening of Birch Town — a European magical town where every fairy-tale resident already lives, works and bargains. The camera never leaves first person: a 26-year-old Chinese man wakes in the attic bedroom above his pawn shop and finds Xuexue, an adult Snow White, leaning over him.

Five beats, four hard cuts: the sleepy close-up POV; the startled sit-up with both real hands closing gently on her shoulders; the teasing morning exchange; the slow glance down her exact costume and back; and the final beat, where she unfolds an aged parchment and asks for the Truth Nail before the eight o’clock hearing. Hybrid media throughout — a live-action room and a real male POV body, composited with one premium photorealistic 3D CG character. No male face, ever; no third-person cut; no music, only dialogue and room sound.

Watch it now: Episode 1 · Visitor at the Bedside · 9:16 Vertical →

30 seconds · 9:16 vertical · 1080×1920 · Chinese dialogue, no music — rendered with Seedance 2.5 (image-to-video) in a single continuous pass from five strict reference images, with all five hard-cut beats and the off-camera dialogue authored directly in the prompt.

Download the MP4 →

“30 seconds, vertical 9:16, 1080p. Hybrid-media live-action first-person scene in a real European magical pawn-shop attic bedroom at early morning. Real photographic room and real adult Chinese male POV body, combined with one premium photorealistic 3D CG adult woman named Xuexue. Neutral slightly warm grade, medium-low saturation, natural brightness, medium contrast, compressed highlights, natural skin, real environmental light, shadows and photographic depth of field. No music. VERTICAL COMPOSITION: compose specifically for 9:16, never stretch or squeeze the 16:9 references. In close shots center Xuexue's eyes, face and shoulders in the upper two-thirds while the real male hands enter from the lower edge. In wider shots preserve the vertical depth of the bed, floating ledger, central town window and fireplace; the far-right desk may remain partly outside frame. The costume inspection and parchment reveal use the vertical frame height naturally. ALL FIVE ATTACHED IMAGES ARE STRICT REFERENCES: Reference image 1 is the opening composition anchor. Recreate its first-person lying-down angle, Xuexue directly above the lens at very close distance, head tilt, shoulder angle, warm morning light and depth of field, reframed vertically around her face without changing her identity. Reference image 2 is Xuexue's immutable character sheet. Preserve exactly the same premium photorealistic 3D CG adult female face, proportions, moist dark eyes, pale rendered skin, short wavy black bob, red bow and gold ornament, earrings, choker, navy off-shoulder bodice, ivory ruffled neckline, red ribbons, gold trim, layered ivory-and-navy short skirt, long red back ribbons, dark shoes and 3D rendering style. No face or costume change. She is never a live-action actress. Reference image 3 is the immutable live-action male body and clothing. Preserve the same real 26-year-old Chinese male hands, forearms, skin tone, veins, anatomy, charcoal-black loose sleep shirt with rolled sleeves and matching trousers. The man remains live action, never CG. Reference image 4 is the immutable live-action attic bedroom. Preserve the sloped dark wooden roof, carved bed on the left, rumpled bedding and dark blanket, nightstand and floating black-gold ledger, central multi-pane window with town spire, stone fireplace, dark wardrobe, chandelier, shelves, rug, worn floor and morning light. Do not redesign, mirror, modernize or rearrange it. Reference image 5 is the immutable Truth Nail parchment. Preserve the same aged rolled parchment, central long black-silver nail with domed head, engraved shaft and sharp tip, technical diagrams, Latin manuscript layout, measurements and circular seal. Do not redesign it. No generated relationship image, alignment board or other image may influence the video. CAMERA LOCK: strict first-person head-mounted POV at the real male protagonist's eye height. Camera direction equals his natural gaze. His real hands, forearms and partial clothed torso may enter frame; his head, face, eyes and facial features never appear, including reflections. Natural live-action head inertia: sleepy focus, sudden sit-up, subtle settling, one downward inspection and return to eye contact. No fisheye, no ultra-wide distortion, no gimbal, no floating camera, no third-person shot. MEDIA LOCK: man and bedroom are live-action photography. Xuexue remains visibly premium photorealistic 3D CG with rendered skin detail, wet eyes, individual black hair strands, natural facial-rig movement and physical cloth. She shares the room's light, exposure, cast shadows, contact shadows, depth, occlusion, mattress depression and hand contact. No anime, cartoon, game look, plastic skin, clipping or floating contact. 0.0-4.5 seconds — SHOT 1. Vertical strict first-person view from the man lying in bed, matching reference image 1. Vision starts dim and soft, blinks once, then focuses. Xuexue's face is directly above and close, centered vertically; she bends a little closer, narrows her eyes, smiles, glances at the viewer's lips, then looks into the lens. No spoken words, no dialogue, no narration. Close breathing, morning town ambience, soft mattress creak. Hard cut. 4.5-8.5 seconds — SHOT 2. Xuexue leans closer as if preparing to kiss. The man suddenly wakes and sits upright with real inertia. Both real male hands enter from the lower frame, close completely and gently around her shoulders, then push her back to a stable seated distance. Correct pressure, cloth deformation and contact shadow. Xuexue's eyes widen, then one eyebrow lifts and her smile returns. The sleepy startled male says exactly, “卧槽……百雪宫主？” Bed shake, clothing friction and quick breath. Hard cut. 8.5-14.5 seconds — SHOT 3. The man sits in bed, vertical first-person view facing Xuexue's upper body. His hands release and withdraw. Xuexue settles, straightens her neckline once, smooths her skirt once, leans slightly closer, tilts her head and looks into the lens. She says exactly, “早啊，周老板。想我没？” After she finishes, the man asks exactly, “不是……咱俩之前认识吗？” Quiet street ambience and mattress sound. No overlapping dialogue. Hard cut. 14.5-20.5 seconds — SHOT 4. The man's gaze travels naturally downward along Xuexue's exact costume, using the full vertical frame from bodice to layered skirt, then rises back to her face as he leans against the headboard. He mutters exactly, “不过……长大后的百雪还真挺漂亮。” Xuexue smiles, raises her chin, maintains eye contact and replies exactly, “现在认识也不晚。” One short natural laugh, no kiss and no extra touch. Hard cut. 20.5-30.0 seconds — SHOT 5. The man's real right hand briefly rubs his face without revealing it, then lowers. He asks exactly, “行了，说正事。这么早找我干嘛？” Xuexue's smile fades and her brows draw together. She stands beside the bed. She takes the exact folded parchment from inside the upper edge of her clothing, unfolds it completely with both hands and presents it vertically toward the lens. The parchment becomes the clear foreground subject and exactly matches reference image 5. Xuexue looks over its upper edge and says slowly and exactly, “我要一枚真言镜钉。” Hold on the complete parchment and her serious eyes for the final beat. Crisp paper and fabric movement. VOICE LOCK: the male is a normal 26-year-old Chinese man, slightly rough and not especially pleasant, with a mild natural regional accent and imperfect standard Mandarin. Casual daily speech like someone just waking. No broadcaster diction, dubbing performance, perfect pronunciation, theatrical emotion, vocal fry or deliberately attractive voice. Xuexue has a natural adult female voice, lightly teasing in shots 3-4, slow and serious in shot 5; no anime or childish voice. Accurate Chinese lip movement for Xuexue. Male voice remains off-camera. Separate speaking windows, no overlap. SOUND: close breathing, morning street ambience, mattress and bedding, clothing friction, parchment unfolding and one distant bell. Dialogue and sound effects only, no background music. HARD NEGATIVES: no male head, face, eyes or reflection; no third-person view; no live-action Xuexue; no CG male; no identity drift; no costume, room or parchment change; no extra person; no duplicate Xuexue; no extra limbs, fused fingers, clipping or floating contact; no subtitles, captions or invented text; no fisheye, ultra-wide distortion, beauty filter or plastic skin.”

Model: Seedance 2.5 — image-to-video, 30s @ 1080×1920, five reference images, single pass.

Professional — the cloud harness runs the whole pipeline, from a one-line world to a finished vertical film, with no local setup. Stable access to the latest models: GPT Image 2 Seedance 2.5 Kling Image-To-Video Gemini 3.1 Flash Image.

One-click generation — officially tuned skills drive every step: world building, key art, strict reference frames, image-to-video generation and instant publishing to a shareable link.

Community & IP-first — adapt any world or start fresh from a single sentence. Keep your frames, prompts and sprites, and remix them into new vertical episodes anytime.